← All questions
AI productstretch
Your company wants an AI assistant that answers employee questions from all internal documents. What could go wrong, and how would you prevent it?
How to approach it
Sort the risks before listing them. Showing a document to someone who shouldn't see it usually matters more than an occasional wrong answer, so start there.
How a strong answer opens
“The biggest risk I see isn't a wrong answer, it's a right answer shown to the wrong person. If the assistant can read every document, it could surface salaries or legal files to anyone who asks.”
Outline of a strong answer
Try answering out loud for a few minutes before you open this.
- Sort risks by likelihood and harm: data leaks, wrong answers, hidden instructions, over-reliance
- Make retrieval respect each employee's existing document permissions
- Keep sensitive sources, such as HR and legal files, out of scope at launch
- Defend against instructions hidden inside documents (prompt injection)
- Show sources with every answer so employees can check them
- Red-team before launch, then monitor flagged answers and access patterns
Where this question usually goes wrong
Focusing only on wrong answers and forgetting that retrieval must respect document permissions and that documents can carry hidden instructions.
Learn the method
Risk, safety and trust in Module 7, AI product interviews.
The full kit has worked answers in full, start to finish, for every round type. $129 once.
More ai product questions
- How would you set up a human review process for an AI system that flags fraudulent transactions?
- Design an AI tutor for high school math students.
- How would you reduce hallucinations in an AI feature that answers questions from a company's help docs?
- How would you use AI to improve Google Photos?
- What makes a good eval set, and how would you build one for an AI travel-planning assistant?